Achieving Near-Zero Added Latency for Modern Any Point in Time Vm Replication

2. The method as recited in claim 1, wherein writing the pointer and IO metadata to the splitter journal site does not increase a latency associated with the operations between the application and storage.

3. The method as recited in claim 1, further comprising, asynchronous with operations occurring along an IO path between the application and storage, sending the IO data from the IO buffer to the replication site.

4. The method as recited in claim 1, further comprising maintaining write order fidelity of incoming IOs from the VM as the VM migrates from a first host to a second host, and maintaining write order fidelity comprises marking each incoming IO with a session number.

5. The method as recited in claim 1, further comprising receiving IOs from two different hosts as the VM migrates from one of the hosts to the other host, and maintaining write order fidelity of the IOs.

6. The method as recited in claim 1, further comprising experiencing a crash of the VM and, after restart of the VM, resuming evacuation of the splitter journal at a point where evacuation had previously ceased due to the crash of the VM.

8. The method as recited in claim 1, wherein the memory comprises non-volatile memory (NVM).

9. The method as recited in claim 1, wherein part of the method is performed inside a hypervisor kernel.

10. The method as recited in claim 1, wherein the IO path comprises a path between the application and a splitter, and a path between the splitter and the storage.

12. The non-transitory storage medium as recited in claim 11, wherein writing the pointer and IO metadata to the splitter journal site does not increase a latency associated with the operations between the application and storage.

13. The non-transitory storage medium as recited in claim 11, wherein the operations further comprise, asynchronous with operations occurring along an IO path between the application and storage, sending the IO data from the IO buffer to the replication site.

14. The non-transitory storage medium as recited in claim 11, wherein the operations further comprise maintaining write order fidelity of incoming IOs from the VM as the VM migrates from a first host to a second host, and maintaining write order fidelity comprises marking each incoming IO with a session number.

15. The non-transitory storage medium as recited in claim 11, wherein the operations further comprise receiving IOs from two different hosts as the VM migrates from one of the hosts to the other host, and maintaining write order fidelity of the IOs.

16. The non-transitory storage medium as recited in claim 11, wherein the operations further comprise experiencing a crash of the VM and, after restart of the VM, resuming evacuation of the splitter journal at a point where evacuation had previously ceased due to the crash of the VM.

18. The non-transitory storage medium as recited in claim 11, wherein the memory comprises non-volatile memory (NVM).

19. The non-transitory storage medium as recited in claim 11, wherein one or more of the operations are performed inside a hypervisor kernel.

20. The non-transitory storage medium as recited in claim 11, wherein the IO path comprises a path between the application and a splitter, and a path between the splitter and the storage.